If achievements should be abandoned because they're counter to good design practices, what should take their place? Variants, Keith Burgun argues, and in this article he explains what they are and are not.

At the end of my previous article, An Alternative to Achievements, I suggested that a good replacement for achievements, should we begin to turn away from them, would be variants. What I didn't do was go into great detail about variants, and exactly how we might -- or should -- go about employing them.

What Is a Variant?

First, I should establish clearly what I consider a "variant." People have been making variations of games for as long as there have been games to make variations of, and the line between "what is a variant" and "what is actually a completely new game" can be very blurry. If I turn off the items in Super Smash Bros., is that " Super Smash Bros.: No Items Variant"? How about if I turn off the music in Street Fighter?

While I'm aware that colloquially, neither of those things would likely be considered a variant, for the purposes of this article, I am considering anything that changes the game rules a "variant". So that means that Smash Brothers: No Items is a Smash Brothers variant, but Street Fighter With No Music wouldn't be. I do, in general, believe that by changing a single rule, you are in fact creating a new game. After all, as anyone who has ever tried to balance a game knows, changing just one rule will usually have ripple-effects throughout a game system, often dramatically changing its character.

So a variant is any set of game rules as played, whether those rules manifested in a mod, in a changed setting, in house rules or otherwise. So, for now, imagine if you will that each version of a game where the rules have been changed even a little bit -- each variant -- is its own shiny, brand-new game.

Here are two major points or questions to consider as we move on:

Obviously, it's not fair if the player can change what his objectives are on-the-fly during play. This needs to be agreed upon before play begins in order to be fair. 

One last thing to mention before I proceed is that this article is written for designers, not players. So, even though house rules certainly qualify as a method of creating variants, there won't be much talk of them.

Part 1: Official Variants

Often times, the developer of a given game will come up with a couple of different ways that their system can be used. We can refer to these as "official" variants. In the very early days of digital games, it was almost assumed that a game would be packaged with a number of different ways to play. The super-successful Atari 2600 had a physical switch, right on the face of the console, dedicated entirely to changing "game modes", as shown below.

This would sometimes change minor features, but sometimes switch you to a completely different game altogether, as in the case of the deathmatch shooter game Combat, wherein switching the switch would take you from "tanks that shoot bouncy bullets at each other in a maze" to "dogfighting with a cluster of ships versus a single large ship."

Earlier NES games also often had a similarly vague "mode-selection" mechanism on the title screen, details of which were included only in the game's manual.

Modern games still often come with various modes of play that are accessible from the title screen, whether they be some kind of puzzle-mode, a multiplayer mode, etc. My own 100 Rogues has a number of such modes, partially because we thought they were good ideas, but also maybe a bit because we felt we "should" have a lot of modes. For a long time, the thinking has been, "the more options and ways to play, the better!" While it's easy to understand why this would be thought to be true, there are a couple of real dangers to this sort of "spray and pray" approach.

The first problem is the question of "which is the real game?" That might sound like a silly question with an answer that can only be arbitrary and meaningless, but it's not. What I mean by "real game" is, which is the game that the most intense effort in game design has been put towards?

If the answer is "all of them equally", then that itself is a problem. A great game, as we'll all happily agree, I'm sure, is a rare thing, whereas mediocre or even bad games are extremely common and easy to design. The process of designing a game is a careful one that takes months or years, after which you're still not guaranteed to have a great game idea.

Quantity

A game-release represents an opportunity for a developer to show a consumer a game. Even with that opportunity, many consumers will still never end up playing a given game. The developer also has X hours to spend working on designing (and otherwise creating) that game.

With all that in mind, the idea that the designer has split time evenly between a handful of different games -- even if they're very similar "variants" and not completely new games -- is foolish. It amounts to using a similar model as those shady "100 Games in 1" bundles that you'll sometimes find in mall kiosks. 

In our world, a world saturated with games everywhere we look, one that's becoming more saturated every year, a strategy of quantity over quality is madness!

This, of course, isn't a call to avoid variants altogether. Subtle variants can be very low-cost, and have the benefit of sort of "hedging our bet" a tiny bit in terms of whether or not people will like our title. It's just worth keeping in mind that quality is the goal, not quantity.

Game Options

Modern games also tend to have a slew of "gameplay options," and as per our discussion about "what makes a variant," technically, every one of those gameplay setting combinations is a new variant.

I think we've all played some games that had some options that, if checked, would make the game horrible. Some games allow you to turn off critical gameplay features, ruin balance or pacing, or turn on distracting or noisy features. Arguments rage on about Super Smash Bros. and whether to have items on or off, but putting the items on Very High and nothing but Invincibility Stars would probably result in a dull experience.

One of my favorite examples of this is the Gamecube title Super Mario Strikers, a Mario-themed soccer game. The game had a quite powerful "tackle" move, but there were no penalty cards as in real soccer. The move was so powerful because it had essentially a 100 percent chance to take another player down for a few seconds, removing the ball from their possession. 

The counterbalance to this move, the thing that made it so that you wouldn't just spam it all the time, was that if you were struck by such a move, you were awarded a random "item" (think Mario Kart style).

Now, I'm not crazy about the idea of random items coming in and spamming up this already-ramped up, tight soccer game, so I look in the options, and hey -- there's an option to turn off items! Great! Except now, the game is completely broken, and you've got to house-rule out the tackle.

That's an extreme case, of course, but there are other cases that I'll get into more in the next section.

Server Settings

Multiplayer games -- and I'm thinking specifically first person shooters, here -- will often have player-run online servers for people to play on. One of the perceived upsides of this model is that server administrators can change many of the game rules. I'm not talking "mods," exactly (I'll get to that in the next section) but smaller rule changes such as changing respawn timers, reducing gravity, or changing the way that weapons work.

While it seems like a positive thing, what we're doing when we allow this is we're allowing lay people -- essentially, just some dude who was willing to pay 20 bucks a month for a server -- to change game rules, and thus, to design the game.

As an example, one of my favorite online FPSes is Team Fortress 2. Most of my history of playing the game was spent with me trying to find a server that not only had good ping, a decent amount of players, but also, I needed a server that was actually running the game Valve designed -- "vanilla" TF2.

Sadly, there was no way to filter this out for the longest time. More recently, they've added some filtering options, but as of the last time I checked, there were still some mods that would go undetected by your filter settings. One rule that's notoriously hard to avoid is that of reduced respawn time, where the amount of time you have to wait after being killed is reduced, often dramatically. Valve, at one point, even had to issue a statement on its blog educating people about why this was a bad idea -- namely, that it made matches more likely to stalemate:

They provide a reward for the team that's doing well, in that if they wipe out a significant amount of the enemy team they're rewarded with a short grace period in which they can achieve objectives. Without them, we found teams felt like they'd been penalized when they cleared the enemy off the last capture point, only to have them all return immediately. 

Their blog post had little effect, in my opinion, but the real problem was that Valve was in the position to have to make an "appeal" to people to not ruin its game in the first place. Most players, and hence most server operators, simply do not understand the sometimes counterintuitive ramifications of a rule change, and so they shouldn't be trusted with this responsibility. If game design was something that any random person with zero experience could do well, then we wouldn't need game designers.

Difficulty Modes

Imagine if, when you bought a new book, it actually came with four books inside it. It's four copies of the same book, but written for three different reading levels -- from very simple to very verbose and descriptive. Am I mistaken for thinking that the book-reading community would roundly reject this terrible idea? It's hard enough to write one good book, and most who read books understand this. They would likely want to know: which of these (if any) is the good one?

Difficulty modes are variants, for our purposes. I sense that, while difficulty modes have always existed, there may be some small but growing percentage of us in the game design community that have started to distrust the concept of difficulty modes.

I mentioned the problem of "which is the real game", and nowhere is this problem more prevalent than with difficulty modes. When I start up a new game, I don't want to play the version wherein they just tuned the numbers arbitrarily downward to make it easier, and I don't want to play the version wherein they did the opposite to make it harder. I want to play the balanced difficulty game that they carefully tuned and designed.

But let's say you work hard to make your different difficulty modes actually different, by changing AI patterns, number of enemies, what types of power-ups appear, etc. In this case, we're really talking about a different game, and so we're back to the "Three-in-One Fun Pack!" problem. It doesn't help anyone to have more mediocre games.

In short, avoid difficulty modes if at all possible. If you're going to have difficulty modes, have a very small number of them (preferably two) and label them clearly. Perhaps you want an "introductory" skill level for players who are just starting; this is understandable. Consider naming that "Easy Mode" and the normal game "Normal Game," for clarity. 

The same thing goes for creating a harder mode -- the important thing is that your players can tell which is your "balanced difficulty", and which is a variant. Sure, plenty of games already have a "normal" difficulty setting, but until developers adopt the position that one of these games is their "real" intended game design, we, the players, have no way to know which of the difficulty modes it is. I think we've all played a game where "Hard" felt like the version that the developers intended (the recent XCOM comes to mind).

Part 2: Mods

It may seem that I advocate for some kind of anti-mod position, but that's not the case at all. Mods provide tremendous value to both the community and the developer, and sometimes the mod can even outshine the game that's being modified. One of the most famous cases of this was with Counter-Strike, which was a mod for Half-Life; another being the Defense of the Ancients mod (DOTA) for Warcraft III. It's arguable that both of these games had a larger, more lasting impact on the world of games than either of their base games did -- which is saying a lot, in this case!

So, obviously mods -- basically, user-made variants of games -- are incredibly important. Developers have known for years that making sure that mods are easy to make is a very important part of game development. I agree with this commonly held wisdom.

In fact, developers can go further with this than they already do by "sponsoring" some of the best mods. A beautiful thing, which has happened a few times before, is when a team makes a mod that's so fantastic and popular, that the development team formally endorses that mod. There have even been cases of these mod teams being hired or becoming a small offshoot of the original company.

But what we haven't seen yet is a kind of institutionalizing of community mod embracing, so much. Some games are starting do this, but not all of those are necessarily doing it in the right way.

Some developers have gotten the idea that it would be cool if the community can create things that actually get regularly added to the base vanilla game. While this is certainly exciting for the users, who get to get their 15 minutes of fame for a bit of content they created, this is a disastrous premise for the games themselves.

Firstly, the premise of "regularly adding" to any game system is problematic. For example, there has never been a game that sought to "regularly add content" to it that held up in any kind of competitive play setting without some kind of "banning" or house-ruling to fix major balance problems. There is some amount of content for any system that is optimal, and "regularly adding" is going to go over that amount, by definition.

But secondly, regularly adding content that was designed by non-game designers is doubly terrible. Like I mentioned earlier, there are plenty of ideas that seem like a good idea until you do the hard work of sitting down and considering all of the ramifications, many of which aren't immediately apparent and some of which are even counter-intuitive. Implementing a new rule (and thus, a new variant) is a full-time job, assuming you want it to be done right.

To cap this part of the section off, I basically just need to re-iterate a few key points:

In the rare case that a mod makes some change that only improves (without dramatically changing the nature of) the base game, consider adding that feature in a patch, and giving credit to the modder for the idea!

Part 3: Customizable Games

Games with asymmetrical forces, such as... well, most modern competitive video games, but most famously games like StarCraft or Street Fighter, can be really interesting because of the fact that the game starts off being asymmetrical, whereas non-asymmetrical games take at least a little bit to become asymmetrical. 

I've written a bit more on asymmetry and some of its dangers in my previous article on gameplay balance, so I'm not going to get into great detail on the subject here. Suffice it to say though, that each pairing of Street Fighter characters is technically its own variant. Therefore, it makes sense to be somewhat conservative about the overall total number of these variants that you're including with your game.

Customization is arguably even more popular these days. Probably the first major horseman of this apocalypse was Magic: The Gathering, the collectible card game. In this game, as well as in other games in genre that it spawned (collectible or "trading" card games), you're allowed to purchase or otherwise obtain some cards of many thousands of different possible types. Out of the cards you own, you create a "deck," and a large part of playing this game well is building that deck intelligently.

Other games that aren't in the CCG genre have followed suit of late. The online Battlefield series allow for customization in terms of what weapon of many you'll select. League of Legends would be another example. Many games also have non-gameplay customization features, such as the ability to wear different outfits and such. 

The big draw of customizable games is the same as that in an asymmetrical game, only magnified. Now, instead of just choosing one character out of 24, you can create a completely unique "character." The concept is that allowing players some vehicle for self-expression is a good thing.

And that's true -- it is a good thing. It's valuable to players that they are able to express themselves in these ways in these games. However, there are many things that would be valuable to players, but which everyone agrees should not be added to a game due to the costs. I argue that for the most part, customization falls into this category.

The problem with customization is that it is reliant on massive amounts of content, which basically means that balance is impossible. It's similar to -- and usually includes -- the problem of "perpetually adding content" to a system. If the system has enough stuff that you can "customize" it based on your personal preferences, then it definitely has way more content than it needs. 

Of course, there seems to be a great deal of demand in the marketplace for this sort of thing, with a game like League of Legends making more money on some of its customization features than it has with anything else. As we all know, though, "what makes for a strong design" and "what makes more money" don't always align. In case it's not already clear, I'm particularly concerned with the former problem.

One caveat would be that in a system where social interaction is more important than gameplay, such as perhaps some types of "social games" or MMOs. In these systems, it might be wise to have some kind of customization features.

Secondly, it's usually okay to have some level of customization so long as this doesn't affect gameplay. For instance, having a differently colored costume for a character probably won't have much of a negative impact on your game. Non-gameplay stuff only starts to matter once your gameplay elements start to become less visually clear, which again reminds me of Team Fortress 2. When this game was released, it was beautiful looking. All the in-game art assets were consistent with each other, and you can listen to the TF2 developer commentary to hear about how they carefully considered stuff like character silhouettes, which is the first way human beings are able to tell objects apart from each other.

As most people know by now, TF2 is now all spammed up with thousands of hats and other weird accessories, not to mention dozens of extra guns, some of which are only model-swaps with no gameplay difference. So, the game is vastly less clear now than it was on day one, and that has turned many players off from the game.

What about a case where "customization" is actually gameplay-essential, as in the case of equipping items in an RPG? My answer would be that if there's some system there which is so un-strategically interesting / flat / trivial, that you can leave the decision to some kind of personal "preference", then it's dead weight on the system. Classic example: do you want +1 attack, or +1 defense? Obviously, you need both; there's no way to use strategy to decide which one to get, so in this case you could simply choose based on your personal whim. The problem is that the system is asking you questions whose answers are actually meaningless.

Part 4: More on the Alternative to Achievements

I mentioned the role that variants could take in replacing achievements, and I'd like to talk a bit more about that here.

Go through a list of achievements for one of your favorite games. Try to isolate the top five or 10 achievements or so. Let's say one says "win a fight without using grenades" or something. Now, instead of this being an always-on thing that gets triggered once you beat the game without using grenades, there's now a new Challenges Menu from the title screen.

On this hypothetical challenges menu, there are a number of, well, basically "achievements" that must be chosen from the start. It doesn't really make sense to have 150 different goals active at once. Beyond being noisy, the player will certainly end up getting some of these unwittingly, by accident, while trying to achieve some other goal.

So, the player can choose the "No Grenades Challenge", and then the game begins. Only once he's told the system that he's going for this challenge can be awarded that "medal" or, if you prefer, "achievement," for having won the game. If he achieves this without having told the system that he's trying to achieve it, it's not worth anything.

This might seem strange, but it's again important to remember that this is how games work! When we start a game, we must know precisely what the goals are, and are not. A player in chess can't capture a your rook, and then declare victory. "I was going for your rook! That was my goal!" The goal must be clear and agreed on by all participants up front, and no, single-player games are no exception to this.

Part 5: Marketing Variants

Variants, unlike in-game content, are a great way to expand your game. There is literally absolutely no harm that can come to your game by selling players another game mode.

Puzzle modes of non-puzzle games are really great, not just for players, but even for designers. Creating a puzzle mode variant, where it's a non-random, pre-built puzzle out of your game system, can be a great way to analyze the kinds of depths that your game actions have. If you can't figure out how to make a clever puzzle using your gameplay's actions, then that's a pretty good sign that your system doesn't have enough depth, because the way that puzzles become hard is through non-obvious use of those actions.

I did this with 100 Rogues in a mode we called Challenge Mode. I found this mode, a puzzle-room mode, to be extremely useful in testing out the player abilities of the different classes. Many of the abilities got leveled-up -- or in at least one case, completely changed -- because of this phase of testing.

Other great modes to have are more story-based campaigns and episodes. It's a great, non-destructive way for people to experience the world you've created in a new way.

These are only a couple of examples, of many, many possibilities. With so much pressure to have good, fair DLC for our games, we really need to start focusing on different kinds of game modes we can add.

At the end of the day, variants simply represent new ways to experience your gameplay concept. Most variants won't be as good as the base game, but most variants also cost far less to "make", being that they are simply a modification of something that already exists, so overall, they're often worth the trouble. As long as we never allow ourselves to get carried away, or to lose the original game concept, exploring the possibilities of variants is only a positive thing.